Stark differences in HIV prevalence and incidence between black and white MSM are well documented, but dozens of studies have shown that the differences in HIV prevalence are not attributable to higher levels of high risk sex, to higher numbers of sexual partners, or to more drug use among black MSM. Therefore, there is an urgent need to conduct studies to examine factors beyond individual-level risk behaviors (e.g., dyadic and network factors such as concurrency, and community-level factors such as neighborhood-level indicators of disease burden and poverty) and their relationship to HIV risk for black MSM. We propose a prospective cohort study of 340 HIV-negative black MSM and 340 white MSM, with baseline ascertainment and periodic re-ascertainment of individual-level, dyadic and network-level, and community-level factors to better describe the likely complex array of factors that give rise to disparities in HIV prevalence and incidence in black and white MSM. MSM will be recruited using a well- described method of venue-time-space sampling; we will administer an interview and HIV, STI, and drug testing at baseline, and then repeat both interview and testing at 3, 6, 12, 18, and 24 months. Primary outcomes will be prevalence of HIV, STI and undiagnosed HIV infection, and incidence of HIV, a combined HIV/STI outcome, and unprotected anal intercourse. Information on network and recent partner characteristics will be collected at baseline and follow-up visits. Analyses will examine individual, dyadic and network, and community level factors associated with prevalent or incident outcomes. Our questionnaire will include theoretically-driven measures based on an ecological systems model, which will allow us to translate findings from our study into the development of future HIV prevention interventions for black and white MSM.